ABSTRACT Natural killer cells are cytotoxic against virally infected cells and cancerous cells, but had never been shown to kill host cells harboring intracellular bacteria in vivo. We discovered the first two examples of this basic function of NK cells, which can kill hepatocytes infected by Listeria monocytogenes or Chromobacterium violaceum. This activity required exogenous or endogenous IL-18, respectively. This discovery originated from our search for bacteria that would fail to evade caspase-1 detection in vivo. A series of logical steps led us to discover that C. violaceum is extremely lethal to caspase-1 deficient mice, which succumb to as few as 100 CFUs. In stark contrast, WT mice are fully resistant, surviving 1,000,000 CFU systemic challenge. This new model led us to make a surprising new discovery. Approximately half of the caspase-1 directed defense was via IL-18, which stimulates NK cells to kill C. violaceum infected hepatocytes via perforin mediated cytotoxicity. Vertebrate adapted pathogens have strong selective pressure to evade inflammasomes ? we have previously show this is the case for L. monocytogenes. This led us to postulate that L. monocytogenes evades NK cells by preventing IL-18 secretion, and indeed, when we treated mice with therapeutic IL-18, the NK cytotoxic response is restored. In this grant we explore how NK cytotoxicity drives the clearance of L. monocytogenes and C. violaceum. We study the role of NKG2D in identifying infected cells, explore the proteins that are targeted by granzymes in the target infected hepatocytes, and study the fate of infected hepatocytes after they are killed.